We will now fit our stocks data of the upper tail beyond the chosen threshold of 0.05 to a GPD method.
## $threshold
## [1] 0.05
##
## $nexc
## [1] 32
##
## $conv
## [1] 0
##
## $nllh
## [1] -74.3385
##
## $mle
## [1] 0.01965532 0.60647468
##
## $rate
## [1] 0.04238411
##
## $se
## [1] 0.00558524 0.24858608
By using the POT approach, we built a model for high values of the negative log-returns where we obtain the Maximum Likelihood Estimates for the scale (sigma) and shape (ksi) parameters/coefficients : 0.01965532 and 0.60647468 [ARE THEY THE COEFF THEY ASKED TO REPORT ??], with Standard Errors being 0.00558524 and 0.24858608. [DOES THIS CORRESPONDS TO THE MEASURE OF UNCERTAINTY ??]
## 5%
## -0.04455341